The canary song system has proven to be a rich source of information on such general issues as neural sexual dimorphism, effects of steroids on brain development and function, adult neural plasticity, recovery of function, and lateralization. The proposed research will integrate behavioral observation, hormone manipulation, and sophisticated quantitative microanatomical techniques in adult canaries to further study these issues. The first two experiments study the behavioral and neuroanatomical consequences of experimentally preventing the normal fall drop in sex steroids. 1. If treatment with testosterone (T) prevents the neural regression and later song reorganization seen in untreated males, it would suggest that processing space in the neural network controlling song is limited: anatomical "erasure" of old circuits is necessary to permit the synaptic realignment which encodes the next year's song. 2. Lesions in left song control nuclei result in substantial song disruption. Recovery using the right hemisphere occurs the following spring. If T-treatment prevents this behavioral recovery, it would suggest that T can maintain lateralized function, and must be removed if there is to be any reorganization of function. Experiment three combines Golgi staining and electron microscopy to find and characterize new synapses in the adult. The P.I.'s computer-microscope system will be used to highlight sites of T-induced dendritic growth. Subsequent EM examination will indicate whether synapses on these new dendrites have a distinctive appearance, and if so, whether new synapses occur elsewhere on neurons from T-treated or control birds. Experiment four combines Golgi staining and autoradiography in order to characterize the song system neurons which derive from adult neurogenesis, and to determine whether endogenous steroids affect their differentiation or fate. Thus, does adult neurogenesis contribute significantly to rejuvenated capacity in the song system? These experiments study questions closely related to human health. Must neural regression occur for there to be forgetting? Are there circumstances under which some behavior must be forgotten in order to permit acquisition of another? Is reallocation of function (a mechanism for recovery from brain damage which also occurs in humans) facilitated by general neural regression in the damaged area? Can the morphology of new synapses and new neurons in an adult nervous system be used to predict conditions which augment their occurance and functional incorporation into a brain region?